Operations Research 2025
Abstract Submission

208. An Electric Ride-Hailing System with Modular Vehicle Convoys for Rural Regions

Invited abstract in session WC-10: Tour Planning Problems, stream Mobility, Transportation, and Traffic.

Wednesday, 13:30-15:00
Room: H16

Authors (first author is the speaker)

1. Philipp Speckenmeyer
Information Systems, Paderborn University
2. Miriam Stumpe
Management Information Systems, Paderborn University
3. Levi Kletetzka
Operations Research, University of Paderborn
4. Guido Schryen
Department of Management Information Systems, Paderborn University

Abstract

Providing sustainable and flexible on-demand mobility solutions in rural areas remains a major challenge due to low population densities and limited public transport options. One promising concept is a modular, on-demand transportation system consisting of small, autonomous electric vehicles that can dynamically couple and decouple, enabling scalable capacities to improve occupancy rates and reduce energy consumption per passenger-kilometer. Such a system is currently being developed in the NeMo.bil project. In this context, we consider the resulting operational planning problem, which extends traditional dynamic ride-hailing with electric vehicles by additionally incorporating decisions on convoy formation. Given a daily demand scenario, operational area, and modular vehicle fleet, we formulate the planning tasks as a combination of vehicle assignment, charging scheduling under range constraints, and convoy formation decisions. To cope with the complexity of real-time, dynamic request arrivals, we model the system as a Sequential Decision Process (SDP) and develop a policy-based solution approach that enables rapid operational decision-making. Our policy aims to optimize system efficiency — such as minimizing total fleet travel distance — while ensuring high service quality, including low request rejection rates. We evaluate our approach by integrating the decision policy into the SUMO simulation framework, using demand data derived from a real-world on-demand shuttle service operating in Höxter, Germany. We compare the performance of our method with that of the existing system.

Keywords

Status: accepted

Back to the list of papers